Byetta's data definitely look better. AMLN reported increased benign thyroid c-cell adenomas and no carcinomas that did not affect survival and were found at concentrations as high as 90 times the human dose.
Think AMLN can get away with this c-cell growth issue. My worry regarding AMLN (as I've posted on SI, Biotech Valuation board) is that LAR's data package is made up mostly of sigle ~300 patient, open label study, based on the FDA classifying LAR as a line extension of Byetta. I see a possibility that the FDA might change its mind, say this isn't enough and ask for more data from the DURATION trials before approving. There's also the FDA's hurdle of comparability between Byetta (commercial vs. trial scale) manufactured at different facilities, and the pancreatitis issue, and the needle size.